摘要:In a feasibility trial comparing two forms of combined inhibitory control training and goal planning (i.e. food-specific and general) among patients with bulimia nervosa (BN) and binge eating disorder (BED), we found evidence of symptomatic benefit, with stronger effects among participants receiving a food-specific intervention. The aim of the present study was to examine changes in behavioural outcomes and event-related potentials (ERPs; N2 (00) and P3 (00) amplitudes) from baseline to post-intervention that might suggest the mechanisms underpinning these effects. Fifty-five participants completed go/no-go tasks during two electroencephalography (EEG) sessions, at baseline and post-intervention. The go/no-go task included 'go' cues to low energy-dense foods and non-foods, and 'no-go' cues to high energy-dense foods and non-foods. Datasets with poor signal quality and/or outliers were excluded, leaving forty-eight participants (N = 24 BN; N = 24 BED) in the analyses. Participants allocated to the food-specific, compared to the general intervention group, showed significantly greater reductions in reaction time to low energy-dense foods, compared to non-foods, by post-intervention. Commission errors significantly increased from baseline to post-intervention, regardless of stimulus type (food vs. non-food) and intervention group (food-specific vs. general). There were no significant changes in omission errors. P3 (00) amplitudes to 'no-go' cues marginally decreased by post-intervention, but there was no significant interaction with stimulus type (high energy-dense food vs. non-food) or intervention group (food-specific vs. general). There were no significant changes in N2 (00) amplitudes to 'no-go' cues, N2 (00) amplitudes to 'go' cues, or P3 (00) amplitudes to 'go' cues from baseline to post-intervention. Training effects were only marginally captured by these event-related potentials. Limitations and recommendations for future research were discussed.